Impact assessment of a railway noise in an alpine valley

Large part of the existing European legislation on railway noise impact assessment refers to indicators based on long term weighted averages of acoustic intensities (Ldn, Lden, Leq). However, several studies have pointed out that noise event indicators (Lmax, SEL) are more appropriate in order to evaluate annoyance due to intermittent noise sources, such as those due to rail traffic. The computation of these short-term averaged indicators requires the estimation of the instantaneous sound level in the surrounding areas, over a region of several square kilometres. To that purpose we have developed an analytical model, named TR-Noise. TR-Noise adopts a quasi-steady description of outdoor noise propagation to compute cartography of instantaneous sound level and hence the indicators averaged on any timeframe, including short-term. The noise sources are modelled as a weighted sum of monopole and dipole moving at steady velocity along a rail track. The model of sound propagation is based on the ISO/DIS 9613-2 International Standards and takes into account different mechanisms of sound attenuation in the ambient. The model has been used to study the impact assessment of a new railway in a dense populated alpine valley in Italy. Firstly, the model was calibrated by means of a field measurement campaign. Sound level measurements performed close to the rail track have been used to estimate the intensity of the source strength related to different kind of trains. Measurements at larger distances from the rail track allowed us to evaluate the performances of the model and to define the influence of the different factors contributing to sound attenuation. Finally, the model has been used to simulate sound level cartographies due to future scenarios of rail traffic and evaluate their impact on the population.